BIOCHEMISTRY
Purification and Properties of Bovine Prothrombin" Joanne S. Ingwallt and Harold A. Scheragat
A rapid, simple, and reproducible method for the purification of bovine prothrombin, starting with a commercial preparation, is reported. By carrying out isoelectric precipitation, chromatography on DEAESephadex, and ultracentrifugation, a product in 18 yield, free of contaminants as evaluated by the sensitive criterion of disc gel electrophoresis, was obtained. Chemical and physiocochemical characterization studies were performed to determine the size and shape of the molecule. Prothrombin in phosphate buffer has a sedimentation equilibrium molecular weight of 74,000 4100, a sedimentation coefficient of 4.80 S, intrinsic viscosity of 3.4 cc/g, a value for p of 2.07 X lo6, posABSTRACT:
z
*
T
he critical step in the blood clotting process is the formation of the fibrin clot by the action of the enzyme thrombin on fibrinogen. Generation of thrombin from its inactive plasma precursor prothrombin has been the subject of much investigation for decades. At present there are two conflicting schools of thought regarding the role which the molecule prothrombin plays in the blood clotting scheme. One of them, led by Seegers and coworkers, suggests that prothrombin is able to generate thrombin autocatalytically in a suitable sodium citrate (Seegers, environment such as 25 1965). In addition, one of several prothrombin derivatives, depending upon the accelerators supplied to the system, is also generated. One of these additional products is an ehzyme (autoprothrombin C); the others are reported to be derivatives of the parent prothrombin molecule and can be activated to thrombin under certain conditions. In this view, prothrombin is a multienzyme system that is able to produce several different product enzymes, depending upon the system used to accelerate the activation of prothrombin. The other approach, suggested independently by Macfarlane (1964) and by Davie and Ratnoff (1964), postulates that the traditional clotting factors present in plasma (factors XII, XI, IX, VIII, X, V, and phospholipid from the platelets) participate in a cascade of proenzyme+nzyme conversions leading to the formation of an enzyme or enzyme complex which is capable of acti-
z
1860
* From the Department of Chemistry, Cornel1 University, Ithaca, New York 14850. Received October 17, 1968. This investigation was supported by a grant (HE-01662) from the National Heart Institute, U. S. Public Health Service. t Predoctoral fellow of the National Institute of General Medical Sciences of the National Institutes of Health, 19651968. $. To whom requests for reprints should be addressed.
INGWALL A N D
SCHERAGA
sesses low helix content, and contains 2 . 3 x carbohydrate as hexose. In the denaturing solvents 6 M guanidine hydrochloride and 6 M guanidine hydrochloride plus 0.1 M 0-mercaptoethanol, prothrombin has essentially the same molecular weight as in buffer, indicating that the protein molecule consists of a single chain. Values for s and [q] in these solvents are also consistent with those for proteins known to be made up of one chain in these solvents. We conclude that prothrombin is a globular compact protein molecule of one chain, possessing little helix content. Preliminary activation. studies are also reported, and indicate that pure prothrombin does not activate autocatalytically.
vating prothrombin to thrombin. The prothrombin derivatives of Seegers are identified with these traditional clotting factors (e.g., autoprothrombin C is reported to be the same as activated factor x ; Spaet, 1964). Proponents of the cascade theory, therefore, suggest that activation of prothrombin in citrate is caused by the presence of contaminating clotting factors in the prothrombin preparations. The conflicts and the similarities involved in these two approaches have been summarized (Davie and Ratnoff, 1965; Kline, 1965; Macfarlane, 1966,1968). It is our view that resolution of these conflicts will be possible only when each of the blood clotting factors is obtained free from other contaminating clotting factors and is chemically and physicochemically well characterized (Scheraga, 1962). Because of the prominent position of prothrombin in the clotting scheme, many workers have attempted to purify and characterize prothrombin from several species, such as man (Lanchantin et a[., 1963), cow (Moore et al., 1965;' Malhotra and Carter, 1968; Magnusson, 1965b), horse (Miller and Phelan, 1967), and rat (Li and Olson, 1967). In general, these different groups report conflicting results on the characteristics of prothrombin. Some preparations have not been demonstrated to be pure, and/or physicochemical characterization has not been reported. In general, then, these studies have achieved only partial success. This project was undertaken several years ago to develop a reproducible purification method for bovine prothrombin and to characterize this product by chemical and physicochemical techniques, in preparation for well-controlled activation studies. The starting material was a partially purified sample of prothrombin pre-
1 We are indebted to Dr. John R. Carter for sending us a copy of this paper prior to publication.
VOL.
8,
NO.
5,
MAY
1969
pared by the Sigma Chemical Co. (St. Louis, Mo.) using a method essentially the same as that of Moore et al. (1965). The Sigma material was then subjected to isoelectric precipitation and chromatography on DEAESephadex, A-50, to yield a prothrombin product free from any contaminants, as determined by polyacrylamide disc gel electrophoresis. This material was characterized as to molecular weight, sedimentation coefficient, and intrinsic viscosity in phosphate buffer and in denaturing solvents, optical rotatory dispersion, amino acid analysis, and per cent carbohydrate as hexose. Preliminary activation studies are also reported. During the preparation of this manuscript, the work of Tishkoff et a/. (1968) on the purification and characterization of bovine prothrombin, using techniques similar to ours, was published. A comparison of the results presented here and those obtained by Tishkoff et al. (1968) is included in the Discussion. Experimental Procedures Purification of Bocine Prothrombin. Bovine prothrombin was prepared using a modification and extension of the method described by Moore et al. (1965). Sigma Chemical Co. performed the initial steps in the purification procedure involving large volumes. These steps consisted of the adsorption of prothrombin from plasma onto barium citrate, elution of prothrombin from the barium citrate-prothrombin complex with EDTA, dialysis of the prothrombin solution against deionized distilled water, and ammonium sulfate fractionation (retaining the 0.5-0.67 M fraction). The prothrombin was then dissolved in 0.05 M phosphate buffer (pH 7.0) and frozen for shipment. Approximately 1 1. of frozen prothrombin solution, containing -1 X lo6 p-tosyl-L-arginine methyl ester units (Ehrenpreis and Scheraga, 1957), from each of three preparations (lots 86B-8550, 87B-8010, and 78B-8200), was received. The solution was thawed, divided into portions of 100-150 ml, and stored at - 70". The aliquots of the Sigma preparations were thawed and dialyzed with frequent changes of the dialysate against deionized distilled water for 12-36 hr at 2-4" in cellophane dialysis bags. The dialysis tubing had been treated at 70" for 15 min in EDTA, rinsed, and stored in the cold in deionized distilled water. After dialysis, the pH of the solution (-pH 6.8) was adjusted to 5.4 with dilute HC1. A precipitate was removed by centrifugation at 7500 rpm for 30 min at 4-6". The pH of the supernatant was lowered to 4.1-4.2 with 0.25 HCI, usually in steps of 0.4 pH unit, for isoelectric precipitation of prothrombin. The prothrombin precipitates which formed were collected by centrifugation at 3500 rpm for 20 min at 4-6", and then dissolved in 0.1 M phosphate buffer (pH 6.0). The solution was usually clarified by ultracentrifugation at 30,000 rpm for 5 hr at 4-6" and then concentrated to 2-10 ml (3-13 mg/ ml) in a Diaflo Model 50 ultrafiltration cell using UM-1 membranes at 2-4". The above solution was then chromatographed on DEAE-Sephadex, A-50 fine (either in the phosphate or chloride ion form ; Pharmacia, Uppsala, Sweden), at
2-4". The Sephadex had been equilibrated with 0.1 M phosphate buffer (pH 6.0) at 2-4" before application of the prothrombin solution. A linear salt gradient (0.01.0 M NaCl in 0.1 M phosphate buffer, pH 6.0) was used to elute the protein. Flow rates of 6-10 ml/hr were obtained for 3.5 X 5-10 cm columns. Fractions (1.5 ml) were collected using a Technicon time-flow automatic fraction collector, and the eluted protein was detected by reading the optical density of the fractions at 280 mp. All fractions common to the prothrombin peak were pooled and concentrated to -3.3 mg/ml on the Diaflo apparatus. The extruded buffer from the Diaflo unit was used in subsequent experiments to dilute the stock prothrombin solutions and served as a reference solvent. Occasionally, ultracentrifugation at 30,000 rpm for 5 hr at 4-6" was performed after the chromatography. Stock solutions of prothrombin were stored at 4" and used within 1 week. Assay for Prothrombin. The method used to assay for prothrombin was essentially that described by Ware and Seegers (1949). Prothrombin solutions (200 pl; optical density of 0.3-1.5) was incubated in 2.3 ml of Bacto Ac-Globulin (Difco, Detroit, Mich.) for 30 rnin at 28"; 1.0 ml of this solution was then added to 3.0 ml of Bacto-Prothrombin 11-stage reagent (Difco) to begin the incubation. At regular intervals over a time span of 60-80 min, 0.4-ml aliquots of the above incubation mixture were added to 0.4 ml of fibrinogen solution (see below), and the time required for the appearance of a visible clot was measured. These assays were continued until the clotting time became independent of the incubation time. The Bacto products were dissolved according to manufacturer's directions. The fibrinogen solution was prepared as follows. Armour bovine fibrinogen (1.O g) (lot U4704, containing -5OZ citrate) was stirred into 50 ml of 5.0% imidazole buffer (pH 7.2) for 15 min. The solution was filtered and allowed to stand for 30 min. The solution became cloudy and was again filtered. The optical density of fibrinogen solutions prepared in this manner did not change for 60-80 min, indicating that these solutions were stable throughout the clotting assay. The solution was approximately 0.35 % fibrinogen, calculated using the factor 0.625 mg/ml per optical density unit at 280 mp (Ehrenpreis and Scheraga, 1957). The number of clotting units per milliliter detected in the assay was taken to be t c - I X lo2 X dilution factor of assay, where t, is the clotting time in seconds. Similarly, the specific activity was set equal to the inverse clotting time X l o 2divided by the optical density of the solution assayed. Actication of Prothrombin. A preliminary study of the conversion of prothrombin to thrombin was carried out by incubating 100 pl of chromatographically purified prothrombin (2.74 mg/ml) separately with each of the following reported activators for 48 hr at 28": 0.1 M phosphate buffer (pH 6.0) as a control; 25.0% sodium citrate ; Bacto-Prothrombin 11-stage reagent (Difco) ; Bacto Ac-Globulin (ciz., factor V) (Difco); BactoProthrombin 11-stage reagent plus Bacto Ac-Globulin; activated factor X (factor X*), 2.0 units/ml (Sigma); purified factor V, 2.0 units/ml (Sigma); factors X* V
PURIFICATION
+
A N D
PROPERTIES
OF BOVINE
1861
PROTHROMBIN
BIOCHEMISTRY
1862
(Sigma); and thrombin, 1.41 mg/ml (Upjohn, Kalamazoo, Mich.), respectively. The Bacto-Prothrombin 11-stage reagent and Bacto Ac-Globulin were dissolved as directed by the manufacturer. The assay for the thrombin which was generated in these prothrombin activation studies was carried out by diluting an aliquot of the activation mixture with 0.4 ml of fibrinogen solution. The number of clotting units generated per milligram of prothrombin originally present in the aliquot assayed was computed from the reciprocal of the clotting time. Polyacrylamide Gel Electrophoresis. Disc electrophoresis was performed with a Buchler Polyanalyst with Buchler power supply type 3-1014A. Polyacrylamide gels (7.573 were used according to Buchler formulations for anionic and cationic systems, running pH 9.3 and 4.3, respectively (see Buchler Instruments Instruction Folder). Acrylamide (American Cyanamid, New York, N. Y.) was recrystallized from acetone-benzene at 2-4". Buffalo Black NBR (Naphthol Blue Black) (Allied Chemical, New York, N. Y.) was used as the staining agent, and bromophenol blue and pyronin B as tracker dyes for the anionic and cationic systems, respectively. Samples were applied in 40 sucrose. Densitometer tracings of the gels were made with a Canalco Model E microdensitometer. We are indebted to Professor F. C. Steward for the use of his microdensitometer. Amino Acid Analysis. The amino acid analyses of prothrombin were carried out with a Technicon amino acid analyzer. Lyophilized salt-free samples were hydrolyzed in constant-boiling HC1 (6.0 M) at 110" for 15, 25, and 35 hr in evacuated, sealed ampules. Corrections for destruction of amino acids during hydrolysis were made by extrapolation to zero hydrolysis time; in the cases of Ile and Val, which are released very slowly during hydrolysis, the data were extrapolated to 70-hr hydrolysis time. Tryptophan was determined spectrophotometrically in 0.1 N NaOH as described by Goodwin and Morton (1946). We are indebted to Mr. Hua Tjan for running the Technicon amino acid analyzer. Other Chemical Analyses. Total carbohydrate was determined as hexose by the anthrone method of Scott and Melvin (1953) using anhydrous dextrose as the standard. Nitrogen was determined by the Kjeldahl method as described by Lang (1958). Extinction Coefficients.The extinction coefficient, for purified prothrombin at 280 mp was determined by two methods. In the first, E was calculated from the number of Tyr, Trp, and Cys residues together with their molar extinction coefficients given by Wetlaufer (1962). e was calculated for both the maximum and minimum values for Cys, because separate assays for cysteine and cystine were not carried out. In the second, e was calculated from the amino acid composition and the amount of ammonia liberated on hydrolysis (both corrected to zero time), together with the Kjeldahl nitrogen content. This extinction coefficient was used for the calculations of protein concentrations, where required. Partial Specific Volume. The partial specific volume,
I N G W A lL L A N D
SCHERAGA
8, for prothrombin in water and in buffers was calculated from the amino acid composition and the values for the partial molal volumes of the amino acid residues as discussed by Cohn and Edsall(l943). This value was used to calculate all molecular weights and corrected sedimentation coefficients. As discussed by Castellino and Barker (1968), 8 of proteins in guanidine hydrochloride solutions may be decreased by 0.01 because of binding of guanidine hydrochloride to the protein. Therefore, the molecular weights and sedimentation coefficients of prothrombin in guanidine hydrochloride and guanidine hydrochloride plus P-mercaptoethanol solutions were also calculated. with the value of 8 in water decreased by 0.01 cc/g. Sedimentation Equilibrium. Molecular weights were determined by the high-speed sedimentation equilibrium method described by Yphantis (1964) using a Spinco Model E ultracentrifuge equipped with Rayleigh interference optics and a temperature control unit. An AN-J rotor was used at speeds below 19,000 rpm, and an AN-D rotor at higher speeds. The cell consisted of the Yphantis six-chambered centerpiece and sapphire windows. Fluorocarbon FC-43 was used as base fluid in the chambers where protein was present; 3-mm column heights were used. All experiments were performed at 23". Rotor speeds were chosen so that the concentration of sedimenting material at the meniscus was essentially zero. All runs were allowed to proceed until there was no further increase in fringe number across the cell. Solutions were prepared by diluting stock solutions of protein with the appropriate dialysate. Prothrombin solutions in phosphate buffer were diluted with the extruded buffer from the Diaflo unit, and this was also used as the reference solvent. The prothrombin-guanidine hydrochloride and prothrombin-(guanidine hydrochloride plus P-mercaptoethanol) solutions were prepared in the following way. Guanidine hydrochloride (Eastman) was decolorized with charcoal and recrystallized from methanol-absolute ether. Guanidine hydrochloride solutions (6 M) were made by adding buffer solution, adjusted to pH 7.4, to the solid guanidine hydrochloride. The buffer composition was 0.021 M Tris, 0.01 M NaCI, 0.01 M EDTA, and 0.1 M P-mercaptoethanol (the latter being omitted in some cases) (Ullmann et al., 1968). Concentrated prothrombin solutions in buffer were diluted with the appropriate guanidine hydrochloride solution and dialyzed against that solution at room temperature for 48-72 hr with frequent changes of the dialysate. The final dialysates were used to dilute the prothrombin-guanidine hydrochloride solutions further, and also as reference solvents. The Rayleigh interference patterns were photographed on Kodak spectroscopic plates, type 11-G backed, and analyzed on a Gaertner two-dimensional comparator. After alignment along the y coordinate, any one fringe at the meniscus was selected as the starting point for counting half-fringes, JJ2. The x coordinate at each half-fringe was measured. Since blank runs showed that deviations from linear parallel fringes were less than 0.057 fringe, no corrections for sedimen-
VOL.
8,
NO.
5,
MAY
1969
tation of buffer salts were made. Weight-average mowere determined from the slopes lecular weights, of plots of In J cs. x 2 , according to eq 1, where x is the
A
&u,,
A B
W
-
Jl,
=
2RT
d In J (1 - Op)w2 dX2
E
(1)
c v)
n
A"
.-c?
l,
a
distance from the center of rotation of the rotor to the position of the half-fringe, R is the gas constant, T is the absolute temperature, D is the partial specific volume, p is the solvent density, and w is the angular velocity of the rotor. Sedimentation Velocity. Sedimentation velocities were measured in a Spinco Model E ultracentrifuge using schlieren optics at 23 '. Double-sector cells with quartz windows were used, and Kodak metallographic plates were used for photography and analyzed on a Gaertner two-dimensional comparator. Sedimentation coefficients were calculated from the rate of movement of the peak and corrected to standard conditions, taken to be 23" in the buffer system 0.1 M phosphate buffer plus 0.65 M NaC1. Sedimentation coefficients for prothrombin in guanidine hydrochloride and in guanidine hydrochloride with P-mercaptoethanol were corrected to standard conditions according to eq 2,
where ?lC,/?)std is the ratio of the viscosity of the guanidine hydrochloride solvent to that of standard buffer, and (1 - gp)std/(l - o p ) G , corrects for the effect of guanidine hydrochloride on 0 and differences in the densities of the reference solvents (Schachman, 1959). Optical Rotation. Optical rotatory dispersion measurements were made with a Cary Model 60 spectropolarimeter, using quartz, water-jacketed cells of path lengths 0.01 and 0.10 dm. An Osram xenon arc lamp (450 W) was used as the light source, and measurements were made in the wavelength range 195-500 mp. The temperature in the cell was controlled with flowing water from a Haake constant-temperature bath. The solutions used were prepared by diluting stock solutions of prothrombin with appropriate solvents. The optical rotatory dispersion data were expressed in terms of the reduced mean residue rotation, [m'IA (in deg cm2/dmole), defined by eq 3, where MR W , the (3) mean residue molecular weight, was taken to be 115, is the specific optical rotation calculated from the concentration of protein in solution, and n is the refractive index of the solution. Appropriate refractive indices were taken from Fasman (1963). The parameters a. and bo of the Moffit-Yang equation (Fasman, 1963) [a]A
0 W
K
A
b
origin
D
-rl
Position ofter electrophoresis
b
anode
'fron!
FIGURE 1: Densitometer
traces of patterns obtained from disc electrophoresis of prothrombin (PT) solutions on 7.5 polyacrylamide gels. (A) Sigma PT, (B) isoelectric product, ( C ) "non-PT" peak from DEAE-Sephadex chromatography,(D) PT peak from DEAE-Sephadex chromatography, after preparative ultracentrifugation at 30,000rpm for 5 hr.
were obtained by plotting the quantity [m'JA((X2XO2)/Xo2) against Xo2/(X2 - ho2),using Xo = 212 mp, Viscosity measurements were made using a CannonUbbelohde viscometer at 25.00 + 0.05'. Outflow times for water were approximately 102 sec. Intrinsic viscosities were obtained by extrapolating the reduced viscosity to zero protein concentration. Calculation of P. The hydrodynamic parameter, /3 (Scheraga and Mandelkern, 1953), was computed from the sedimentation coefficient and intrinsic viscosity by means of (5)
p H measurements were made with a Radiometer type TTTla instrument equipped with a scale expander, type PHA 630 Ta. A Radiometer combined electrode, type GK2021C, was used and the system was standardized with Fisher Certified buffer solutions. Preparatice ultracentrijiugations were performed on a Beckman Model L-2 preparative ultracentrifuge using polypropylene centrifuge tubes. Rotors type 19 and 30 were used, depending upon the volume of solution to be centrifuged and upon the desired rotational speed. Rotor type 19 was used for large volumes and for speeds below 19,000 rpm, and rotor type 30 for higher speeds. Optical density measurements were made with a Zeiss spectrophotometer, M4QI1, using 1 .O-cm cells. Phosphate buffer (0.1 M) was prepared by mixing 0.1 M disodium phosphate and 0.1 M monosodium phosphate to the desired pH.
Results Purification. Further purification of Sigma prothrombin by isoelectric precipitation and chromatography on DEAE-Sephadex led to a pure preparation of prothrombin, as indicated by gel electrophoresis patterns. Figure 1 shows the electrophoresis patterns for the starting ma-
PURIFICATION
AND
PROPERTIES
OF BOVINE
1863
PROTHROMBIN
BIOCHEMISTRY
Elution Volume (ml.1
Chromatographicpattern of the isoelectric product on a 3.5 X 10 cm column of DEAE-Sephadex, A-50, using a linear NaCl gradient (0.&1.0 M) in 0.1 M phosphate buffer (pH 6.0) at a flow rate of -6 ml/hr. The arrow indicates the position of the void volume, UO. Approximately 18 mg of isoelectric product was applied in sucrose, and -12 mg of prothrombin was recovered. FIGURE 2:
1864
terial (Figure 1A) and for the products obtained at various stages of purification. It can be seen from Figure 1A that the starting material contained -4OX prothrombin, 3 W x of another component which appeared to be bovine serum albumin, and at least five to six other components present in small amounts. Isoelectric precipitation removed the bovine serum albumin and some of the other impurities (see Figure 1B). Chromatography of this product on DEAE-Sephadex (see Figure 2) resulted in the final purification of prothrombin. The chromatography was developed with a linear gradient of 0.0-1.0 M NaCl in phosphate buffer as described in the Experimental Section. The elution profile shown in Figure 2 was independent of the protein concentration range used (3-13 mg/ml), and the prothrombin was consistently eluted at a NaCl concentration of 0.58-0.72 M. As a result of the chromatography, prothrombin was eluted after the nonprothrombin impurities, and could therefore be separated from them. Figure 1C,D shows the electrophoresis patterns of the nonprothrombin impurities and prothrombin, respectively. From gel electrophoresis, it was observed that a material, too large to enter the separating gel, appeared in the products of both isoelectric precipitation and chromatography. However, this material could be removed by preparative ultracentrifugation at 30,000 rpm for 5 hr either before or after the chromatography step. Figure lD, which shows the electrophoresis pattern for prothrombin which had been eluted from DEAE-Sephadex and centrifuged, indicates that the protein is pure by this criterion. The peak in Figure 1 labeled prothrombin was found to contain prothrombin by means of the assay described in the Experimental Section. This assay was performed on (nondyed) material obtained from polyacrylamide gels in the following way. After the usual required time for electrophoresis had elapsed, the gel was cut into three to five sections, each of which was placed in an electrophoresis tube fitted with a cellophane bag secured with a sleeve of tygon tubing. The material in each section was eluted from the gel electrophoretically, col-
INGWALL A N D SCHERAGA
.. .
Position a f t e r electrophoresis
3: Densitometer traces of patterns obtained from disc electrophoresis of prothrombin solutions incubated for 24 hr at 28' with 25 sodium citrate (A), Bacto Ac-Globulin (B), thrombin (C), and Bacto-Prothrombin 11-stage reagent plus Bacto Ac-Globulin (D) on 7.5 polyacrylamide gels. Solid lines represent mixtures of prothrombin and activator, and dotted lines represent the activator alone. Patterns for the 48-hr incubation mixtures were different in only two respects, uiz., those fast-moving peaks designated by an arrow disappeared and the prothrombin peaks in B and C were diminished.
FIGURE
lected in the cellophane bag, and assayed for prothrombin. On comparison with a duplicate gel dyed and destained in the usual way, it was found that the gel section which yielded prothrombin as detected by the assay was the one which contained the so-called prothrombin band. Furthermore, it was this band, defined by an R F value, which became more and more dominant during the various purification steps. From an analysis of eight preparations, we find that the over-all yield of protein was 18X, and the over-all yield of prothrombin was 45 Z, with an increase in specific activity of a factor of 2.5. Actiuation. Preliminary activation studies of chromatographically purified prothrombin using various activators yielded the results indicated below and in Figure 3. Solid lines in Figure 3 show the densitometer traces of polyacrylamide gel electrophoresis for prothrombin and the activator after 24-hr incubation, and the dotted lines for the activator alone. Densitometer traces for prothrombin incubated in buffer and in 2 5 % sodium citrate were identical (Figure 3A). No change was observed in the patterns obtained at 24and 48-hr incubation times, and no clot formed in the thrombin assay described in the Experimental Section even after 15-min incubation time with fibrinogen (a blank of fibrinogen in buffer did not clot for > 1 5 min). These results indicate that pure prothrombin is not activated in 25 X sodium citrate under these conditions. An example of those activators [Bacto Ac-Globulin, factor X*, factor V, and factors (X* V)] which generated 5 1-2 clotting units/mg is given in Figure 3B for prothrombin incubated with Bacto Ac-Globulin. After 24-hr incubation, several new diffuse bands were observed in addition to the dominant prothrombin band. By 48 hr, however, the band designated by the arrow disappeared and the amount of prothrombin was sig-
+
VOL.
8,
NO.
5,
MAY
1969
nificantly decreased. Densitometer traces of those activators which demonstrated significant thrombin-generating ability, uiz., thrombin (-10 clotting units/mg, Figure 3C), Bacto-Prothrombin 11-stage reagent ( 2215 clotting units/mg), and Bacto-Prothrombin 11-stage reagent Bacto Ac-Globulin ( 2260 clotting units/mg, Figure 3D), all indicated that the prothrombin band was greatly diminished or absent by 24 hr. However, the fast-migrating band designated by the arrow was present at 24 hr and absent at 48 hr, in common with the activators represented by prothrombin with Bacto AcGlobulin. Chemical Analyses. Some properties of purified prothrombin determined from chemical analyses are summarized in Table I. It is to be noted that the values of
+
TABLE I : Amino
Acid Analysis..**
Amino Acid
Moles/ 74,000 g of Protein
Amino Acid
Moles/ 74,000 g of Protein
ASP Thr Ser Glu Pro GlY Ala Val CYS ('/J
67.7 31.7 41.1 83.7 36.9 57.0 38.0 38.4 20.1
Met Ile Leu Tyrc Phe LYS His Arg Trpc
4.5 23.8 43.7 17.9 22.1 34.0 9.7 41.3 17.6
a Based on a molecular weight of 74,000 (see Table 11). The data in this table correspond to 17.4% total nitrogen. 6 The carbohydrate content (as hexose) was 2.3%, the extinction coefficient, at 280 mp, was 16.5 (from Trp, Tyr, and Cys content and from Kjeldahl and amino acid analyses), and the partial specific volume, 5, was 0.721. c Trp was obtained from the Tyr/Trp ratio, which was determined spectrophotometrically in 0.1 N NaOH (Goodwin and Morton, 1946) and found to be 1.02.
calculated from the Trp, Tyr, and Cys content, agree with that obtained from the micro-Kjeldahl nitrogen determination and the per cent total nitrogen determined from the amino acid analysis. Sedimentation Equilibrium. Plots of In J us. x 2 for prothrombin in buffer and in denaturing solvents for both high (1.0-2.0 mg/ml) and low (0.4 mg/ml) concentrations showed linearity throughout the length of the cell. Quantities used to determine the molecular weights of native and denatured prothrombin are summarized cs. optical density at 280 in Table 11. Plots of 105/Avw mp, showing the dependence of apparent molecular weight upon concentration, are illustrated in Figure 4. Figure 4A gives the plot for prothrombin in buffer; Figure 4B,C for prothrombin in 6 M guanidine hydro-
t 1.50-
1.30I
e
e
, e I .o
I
A
2.0 3.0 Optical Density at 280 mp
4: Plots of 105/2v,+ cs. optical density at 280 mp. (A) Prothrombin in 0.1 M phosphate buffer (pH 6.0) plus -0.65 M NaCl(0). (B) prothrombin in 6 M guanidine hydrochloride. A represents values calculated using D = 0.721, A, for ii = 0.711. (C) prothrombin in 6 M guanidine hydrochloride plus 0.1 M P-mercaptoethanol. rn represents values calculated for ii = 0.721; 0, for 0 = 0.711. FIGURE
chloride and in 6 M guanidine hydrochloride plus 0.1 M P-mercaptoethanol, respectively. Filled symbols represent values obtained for the calculated 8, open circles for a value of 5 decreased by 0.01 cc/g as discussed by Castellino and Barker (1968). All lines were determined by a least-squares fit. Sedimentation uelocity experiments are summarized in Table I11 and Figure 5. In all cases, i.e., for both buffer and denaturing solvents, a single symmetrical peak was observed. The value of s for prothrombin in 0.1 M phosphate buffer (pH 6.0) plus -0.65 M NaCl at 23" extrapolated to zero concentration (Figure 5) was found to be 4.80 S. Using 0.1 M phosphate buffer (pH 6.0) plus -0.65 M NaCl at 23" as the standard conditions, sedimentation coefficients for prothrombin in the denaturing solvents were corrected for differences in the viscosities and densities of the solvents and for ij calculated and decreased by 0.01 cc/g according to eq 3. At a concentration of -3.2 mg/ml, s values for native and denatured prothrombin are compared in Table 111. Intrinsic Viscosity and 0. Intrinsic viscosities, [TI, of prothrombin in buffer and in the denaturing solvents are also given in Table 111. Outflow times were reproduced to within 0.3 %. The quantity, 0, computed from eq 5, has the value 2.07 X 106 for prothrombin in buffer, and 1.58 X 106 and 2.34 X 106 for prothrombin in guanidine hydrochloride and in guanidine hydrochloride plus P-mercaptoethanol, respectively. In the denaturing solvents, the extrapolation of s to zero concentration was not carried out. Optical Rotation. Figure 6 shows the optical rotatory dispersion data in the ultraviolet for prothrombin in 0.1 M phosphate buffer (pH 6.0) plus -0.65 M NaCl at 25". There is a trough near 232 mp, a zero value of [m'IXnear 217 mp, a shoulder at 213 mp, and a positive peak near 198 mp. Curves for the data obtained for prothrombin at 75 and 6", and for the solutions restored to 25", were essentially the same as the one shown. However, gel electrophoresis of a sample which had been heated showed a dominant band at the inter-
PURIFICATION
AND
PROPERTIES
OF
BOVINE
1865
PROTHROMBIN
BIOCHEMISTRY
TABLE 11: The Molecular
Weights of Native and Denatured Prothrombin as Determined by Sedimentation Equilibrium.
Buffer System
0.1 M Phosphate Buffer (pH 6.0) Plus -0.65 M NaCl
Solvent density, g/cc 0, cc/g Molecular weight, *vwb
1.020 0.721 74,000 A 4,100.
6 M Guanidine Hydrochloride 1.140 0.721-0.71 1. 82,000 zt 3,20077,300 3,100c~d 1 .lo-1 .04
*
Lqw denatured/,T, native
6 M Guanidine Hydrochloride plus 0.1 M P-Mercaptoethanol 1.141 0.721-0.71 1. 70,000 3,00065,400 f 2,700Crd 0.95-0.88
*
a As discussed in text. b Calculated from eq 1, and extrapolated to zero concentration. c The standard deviation was computed by the usual least-squares procedure. d The range in LVw corresponds to the range in D.
TABLE 111: Sedimentation
Coefficients and Intrinsic Viscosities of Native and Denatured Prothrombin.
Buffer System
0.1 M Phosphate Buffer (pH 6.0) Plus -0.65 M NaCl
6 M Guanidine Hydrochloride
6 M Guanidine Hydrochloride Plus 0.1 M 0-Mercaptoethanol
Concentration, mg/ml Slope of In x us. t, sec-I Viscosity of solvent/ viscosity, standard* (1 - op)std/(l - op) s, expt1,d S s, cor: S
3.2. 1.59 X 10-6 1 .o
3.2 0.479 X 10-5 1.34
3.1 0.361 1.37
1 .o 4.05 4.05
1 ,49-1,40. 1.22 2,44-2.30.
1 .50-1.40. 0.921 1 .88-1,77~
tal, cc/g
3.4
12-14
< 74
x
a This is the highest concentration shown in Figure 5 . b Standard conditions are taken to be values for 0.1 M phosphate buffer (pH 6.0) plus -0.65 M NaCl at 23”. c The range corresponds to the range in 0. d Determined at w = 59,780 rpm and 23 ’. e From eq 2.
face of the stacking and separating gels and a reduced prothrombin band, indicating the presence of aggregates. A sedimentation velocity experiment with this material yielded a value of -21 S for the “aggregate.” Optical rotatory dispersion data for prothrombin in buffer in the wavelength range 275-500 mp were plotted according to eq 4, and yielded values of -35 for bo and -197 for ao. Similar plots for prothrombin heated or cooled gave essentially the same bo; ao varied with changes in temperature and concentration. Discussion
1866
Purification. In this paper, a rapid reproducible purification method, starting with a commercial preparation of bovine prothrombin, has been described. Electrophoretically pure prothrombin, obtained as a result of isoelectric precipitation, DEAE-Sephadex chromatography, and ultracentrifugation of Sigma prothrombin, has been prepared in a yield of -18% protein (or -45 of the prothrombin available in the starting material) and - 4 7 z clotting units. All steps in the puri-
z
INGWALL A N D
SCHERAGA
fication were monitored using both gel electrophoresis and assays for prothrombin. Using these monitoring tools, it was found that, to obtain a final product of maximum purity, without traces of any contaminating bands upon gel electrophoresis, only the purest prothrombin solutions prepared from the isoelectric precipitation could be used. For this reason, only the precipitate which formed as the pH was varied from
